For minorities, housing discrimination limits access to clean air
A new study finds direct evidence that rental property managers effectively screen minority households into parts of a city exposed to greater pollution.
The wave of demonstrations for social justice and equality in the wake of George Floyd’s murder have been a reminder of just how far cities have to go to become truly inclusive — and that any work toward that future must address the structural discrimination that pervades too many aspects of urban life. Sometimes the destructive forms of discrimination are caught on tape, and rouse us to action. But there are less visible instances that happen every day, which are no less destructive.
A new study sheds light on a troubling example of housing discrimination that occurs on a daily basis across American cities. The research, released in a working paper earlier this year (and updated in April), finds that rental property managers in low-pollution parts of a metro area discriminate against inquiries from African American and Latinx households, relative to white households, but that such racial exclusion doesn’t occur in high-pollution areas. This illegal practice effectively segregates minority households into parts of a city exposed to greater air pollution — impeding economic opportunity, restricting housing options, and exacerbating health disparities.
The work comes from economists Peter Christensen and Ignacio Sarmiento-Barbieri of the University of Illinois, and Christopher Timmins of Duke University. The research trio concludes (my emphasis in bold):
For over two decades, researchers have advanced a racial discrimination hypothesis to explain the factors underlying the disparity in exposures to chemical toxics and other harmful pollutants in the United States. However, no prior study has provided an empirical test. This paper presents experimental evidence that racial discrimination constrains the housing choices of minority households with respect to major polluting facilities in the United States. We find that Hispanic/LatinX and African American renters face strong discriminatory constraints when searching for housing that would limit their exposure to emissions from major sources of chemical toxics in the US.
So how does this happen?
The study: A push for direct evidence
The work aimed to better understand why minority households are disproportionately exposed to pollution in U.S. cities. Housing discrimination has long been considered one reason (among many) for this disparity, but direct evidence of it is hard to isolate through traditional data analysis, given the many social and economic factors that account for urban housing patterns. So the researchers designed a controlled study they call “the first experimental evidence” that discrimination plays a direct, causal role in the link between housing choice and pollution exposure.
The study began by identifying a set of 111 U.S. ZIP codes situated within one mile of a major source of pollution, as tracked by the EPA. These ZIP codes spanned a range of U.S. metro areas, from parts of big metros like New York City to parts of smaller ones like Clarksville, Tennessee. From these locations, the researchers focused on a random sample of 19 areas, using a “major” (unnamed) online housing platform to gather all the rental property listings in the area. All told, they included more than 2,900 listings in their study.
Even within a particular ZIP code, not all neighborhoods are equally exposed to pollution. So the researchers used an EPA measure of pollution concentration to categorize whether a given rental property was located in a low-, medium-, or high-exposure area.
Here, the experiment took a key turn. Using previous behavioral studies of racial stereotyping as their guide, the researchers created a set of names intended to be identified as white, Latinx, or African American (both male and female). They then approached each rental listing using these names, inquiring about availability of the unit. Every property got three inquiries: one from a name meant to be identified as white, one as Latinx, and one as African American. The order of the three inquiries varied from place to place to reduce any potential sequencing bias.
As a final step, the researchers tracked how long it took the property to respond to an inquiry, using the response rate to the white name as a baseline.
The findings: Pervasive discrimination
Across the board, the results were ugly.
In areas with low exposure to pollution, minorities had a 59 percent response rate, relative to whites (meaning minority names received only 59 percent as many responses as white names). Put another way, inquiries from minority names were 41 percent less likely to get a response about an available apartment than names designed to be identified as white. In these areas, black identities faced greater discrimination than Latinx identities, with response rates of 45 percent and 78 percent, respectively.
For listings in medium-exposure areas, the combined response rate to minority names was slightly better, at 71 percent, but still not up to par with whites. But for properties in areas with high exposure to pollution, there was no statistical difference between white and non-white name groups. In fact, in high-exposure areas, Latinx identities were more likely than whites to get a response — suggesting a discrimination in favor of Latinx households moving to polluted areas.
Simply put, the study finds direct evidence that rental property managers engage in exclusionary practices that effectively segregate minority households into polluted parts of a city. Based on nothing but name, minorities can face significantly greater barriers than whites to finding a rental unit in a neighborhood with low exposure to air pollution.
As the researchers drilled down further, they found other troubling results:
- Male minority identities faced greater discrimination than female minority names, with black males getting a response rate of just 28 percent in low-exposure areas.
- Follow-up inquiries faced even stronger levels of discrimination than initial inquiries, with a minority response rate of just 38 percent.
- Filtering the response rate by proximity to a toxic facility instead of neighborhood pollution exposure — a related analysis, but not exactly the same — reached the same conclusion, with a combined minority response rate of 66 percent in close-proximity areas.
- Discrimination was stronger in low-exposure areas with low shares of existing minority households than in with areas with more existing minorities, suggesting a greater tendency in places with lots of white households to exclude new minority entrants.
- Discrimination was stronger in low-exposure areas for higher-priced rental properties than lower-priced ones, suggesting a tendency to extend name biases into perceptions of socioeconomic status.
Whatever the measure — exposure to pollution, proximity to a toxic facility, follow-up inquiry behavior, neighborhood diversity, cost of housing — the practice of discrimination against minority names resulted in fewer opportunities for minorities to live in healthier areas.
As the researchers note, the destructive impact here can take different forms. For minority households that aren’t sure about a neighborhood’s air quality measures, the simple fact that they’re more likely to hear back from properties in high-exposure areas means there’s a better chance their home ends up being located in such an area. But for minority households that focus specifically on finding a home in a low-pollution area, the reduced likelihood of hearing back about an inquiry in these areas means that, on average, their housing search will end up costing more time, energy, and money. The harms accrue even to those who overcome the barriers.
From awareness to progress
The study comes with several caveats that bear mention, starting with the fact that it’s still a working paper and has yet to be peer-reviewed. The analysis was limited to properties found on a single rental platform, which raises questions of whether such findings hold true more broadly. And there are surely other factors aside from pollution that contribute to this type of discrimination, since some property managers may not even know their own neighborhood’s pollution exposure levels.
The researchers also note that the inquiring tenant names were intentionally designed to elicit racial identification, which means they might not reflect the general population. On the flipside, by focusing only on written inquiries, the experimental design almost certainly underestimates housing discrimination, since such practices of course can also occur during in-person site visits and other phases of the process.
One more important note: it must be acknowledged that many minority households didn’t need this study — they already knew such practices to be true from first-hand experience. While documenting direct evidence for discriminatory practice is important in its own right, greater awareness of a problem alone isn’t certain to lead to action addressing it.
Normally, I try to end these Sidewalk Talk study stories with some thoughts on how urban innovation can help ease the challenge at hand. Some steps do come to mind. Increasingly popular rental platforms could design ways of discovering and reporting illegal housing practices, ideally working in collaboration with cities. They might even be able to design privacy-preserving systems that only reveal whether or not a prospective tenant meets rental requirements, without providing unnecessary information that could be used to discriminate (though in the absence of strict enforcement, this effort might just nudge discriminatory landlords toward other screening approaches). And of course there’s a broad suite of energy innovations — especially a push toward affordable electrification — that can improve air quality in our cities.
But it’s important to recognize the limits of urban innovation in the face of a challenge so historically and socially entrenched that it leads a property manager to see a rental inquiry and respond, consciously or not, in a different way based on the sequence of letters that make up the prospective tenant’s name — potentially altering the course of that person’s future and the health of that person’s family. That is not a challenge for innovation to tackle. It’s a systemic barrier that must be torn down through personal reflection, policy reform, and societal growth at the broadest levels.
The goal should be to help provide healthier homes for everyone, not just for some. Getting there requires those of us who work to harness innovation for more inclusive cities to ask ourselves at every step of the way: inclusive for whom?
Follow Sidewalk Labs with our weekly newsletter and our podcasts.
